Donor Impact Claim Audit & Discrepancy Escalation
Verify non-profit programmatic results against internal field logs and send a compliance escalation email before grant submission.
Use this template when auditing grant reports, donor updates, or annual impact disclosures against raw monitoring and evaluation data. It flags discrepancies between field logs and narrative statements to ensure absolute donor accountability.
Role: Lead Evaluation and Grant Compliance Auditor
Context
- Draft narrative report: {{grant_report_draft}}
- Primary funder guidelines: {{donor_guidelines}}
- Ground-truth monitoring datasets: {{field_monitoring_data}}
- Institutional funding partner: {{funding_partner}}
- Material discrepancy threshold: {{discrepancy_threshold}}
- Reporting timeframe: {{reporting_period}}
Task
Conduct an exhaustive audit of all beneficiary counts, financial allocations, and programmatic milestones in {{grant_report_draft}} by cross-referencing {{field_monitoring_data}}, and author a formal escalation email detailing variances, compliance risks, and mandatory adjustments prior to submission to {{funding_partner}}.
Method
- Extract every quantitative metric, beneficiary total, unit delivery figure, and geographic completion claim across {{reporting_period}} from {{grant_report_draft}}.
- Map extracted claims against baseline entries, field monitoring logs, and attendance rosters within {{field_monitoring_data}}.
- Compute the exact percentage and unit variance between draft representations and recorded operational data.
- Flag any metric exceeding {{discrepancy_threshold}} or violating reporting stipulations within {{donor_guidelines}}.
- Investigate potential methodological anomalies (e.g., double-counting of recurring beneficiaries, extrapolation errors).
- Formulate precise, audit-resilient replacement figures with methodology notes to prevent institutional debarment or grant clawbacks.
- Compose an email to program directors and the grants team outlining mandatory revisions and signing off on verified components.
Constraints
- MUST flag any variance exceeding {{discrepancy_threshold}} as a critical blocker.
- MUST NOT extrapolate unrecorded field numbers to fill monitoring gaps.
- Language must be objective, legally defensible, and structured to withstand third-party fiscal audits.
- Keep email within 350 to 600 words excluding the variance table.
Output format
An email deliverable structured as follows:
- Subject Line: [AUDIT ALERT] Pre-Submission Metric Verification: {{funding_partner}} - {{reporting_period}}
- Summary of Audit Finding (Pass / Pass with Revisions / Material Discrepancy Block)
- High-Risk Discrepancy Breakdown (Metric | Claimed in Draft | Verified Ground Data | Variance % | Risk Severity)
- Mandatory Redline Directives (Specific sentences to delete or update)
- Recommended Explanatory Disclosure for Funder
Self-review
- Has every quantitative claim been verified against a verifiable cell or log in {{field_monitoring_data}}?
- Are all calculations of variance mathematically accurate?
- Does the email provide concrete redlines rather than vague requests for clarification?
Explicit role, a named task, and discrete steps the model can follow.
Background, inputs and variables the model needs before it starts.
Hard boundaries — what the model must and must not do.
A named, field-level shape for the response.
Ordered work items that force analysis before an answer.
Length and structure that travel across frontier models.
Signal density — instruction weight without padding.
Documented variables so the scaffold adapts to new inputs.
Quality bar, assumptions and behaviour when inputs are thin.
How much real usage the template has behind it.